REPORT ON THE PETROLOGY OF OCEANIC ISLANDS. 17 



which are common, the pleochroism is strongly marked ; the brown tint is darker for 

 the rays parallel to c, it is yellowish brown for those perpendicular to that direction. 

 The larger felspar crystals are generally much altered ; the hemitropic lamellae are 

 scarcely visible. In certain cases these crystals are filled with secondary products, 

 amongst which calcite may be distinguished ; perhaps this mineral is associated with 

 scales of mica, or of quartz, or secondary felspar. The felspathic crystals of secondary 

 formation, which are scattered throughout the mass, are much more decomposed than 

 those of the first generation. They extinguish at rather small angles, which would 

 seem to refer them to oligoclase. These small plagioclase crystals occur rather often in 

 the form of a cross ; probably we have here to do with a twin analogous to that of 

 Baveno. The sections of augite are large and rather rare ; this mineral is here seen 

 with the ordinary characters which it presents in basalt. It is difficult to determine 

 the ground-mass, as it has been invaded by products of decomposition ; calcite has been 

 developed in certain cavities. If we take into consideration its mineralogical com- 

 position, and if we set aside its structure, which is exceptional, this altered rock might 

 be classed with the amphibolic andesites. 



The harbour of St. Vincent is surrounded by a circle of heights formed of eruptive 

 rocks ; at its entrance there are isolated rocks, which may be considered as having been 

 formerly attached to the chain of hillocks terminating at the coast. These rocks 

 are called Bird Island ; they are covered up to high-tide mark by a wide border of 

 calcareous incrustations consisting of coralhnes. We have examined a specimen coming 

 from this islet ; it is a somewhat fibrous lava, which may be classed with the 

 pyroxenites, 1 and is very closely allied to the basalts. It has the appearance of a basaltic 

 rock ; the very elongated crystals of augite visible to the naked eye are ranged parallel 

 to each other. This disposition determines an almost fibrous structure in the rock, all 

 the vesicles being stretched in the direction of the elongation of the pyroxenic crystals. 

 With the microscope it is ascertained that the felspathic element is not present, and 

 that this lava is essentially formed of augite. Some crystals of that species are por- 

 phyritic, as has just been said, others are microbthic. The large crystals of pyroxene 

 present remarkable pecubarities, as is shown by their microscopical examination. They 

 assume a lengthening quite unusual for this species ; they may attain a length of 7 to 

 8 millimetres, with a breadth of O'l mm. On following one of these sections of augite in 

 all its length under polarised light, it is seen that it extinguishes simultaneously between 

 crossed nicols ; it is therefore a single crystal which extends from one end of the section 

 to the other. 



1 Doelter, loc. cit, p. 187. 



(PHYB. CHBM. CHALL. EXP. PART VII. 1889.) 3 



